Hi everyone. I have modified the dock.py script (in eginput/diffTens/dock) to include the PRE terms from newRefine.py (from eginput/pre) such that PRE restraints are used directly to dock two proteins.
After much trial and error, the script is now outputting a complex of my two proteins which are initially derived from input coordinates (PDBs). In my initial calculations, I was using dummy gamma errors (error in the PRE value) that were low. In these initial calculations, the docked proteins reflected the PRE data. I have now started to include the real errors based on my spectra, which are very large. Now, the proteins are either not moving from their initial position, or are coming close together but not in a way that reflects the PRE restraints. Is there a way to weigh these errors? It may be that as a result of significant T2 relaxation, some of the errors are inherently large. Does anyone have experience in dealing with large errors in gamma values from PRE data? Also, from my understanding, PRE is also influenced by the population of the transient complex. Is there a way to directly specify a low population in the script so that the PRE restraints are weighted differently (e.g., 2% v. 5% of the transient complex in existence)? Thanks in advance for any response! Valerie
